Canon, Fuji, Nikon, and Sony Go Head to Head in This Color Science Battle

Color science is something that pops up in camera debates from time to time. We all hear about how amazing Canon's colors are or maybe a little bit of smack talk towards Sony's color rendition, etc. Tony Northrup recently decided to apply the scientific method to the big four camera brands (Canon, Fuji, Nikon, and Sony) to see who really has the best color, and the results honestly surprised me a lot.

In the video, Northrup likens judging color science to wine tasting. No matter how much someone claims to know what they are talking about or how perfect their sense of taste is (or in this case, sight), people will have a hard time differentiating between any two samples in a blind test. Tony prepped this quiz with straight out of camera JPEGs with auto white balance and found that as soon as people thought an image was made by a certain manufacturer, they instantly showed their bias. While I do have a small problem with his methodology (the Nikon Z7 got the white balance super warm and I feel that they should have all been set to their respective daylight presets), it is an extraordinarily fascinating video. What do you think of the results? Were you expecting Canon to end up at the bottom? All in all, it shows just how little color science matters. The difference is minute.

David J. Fulde's picture

David J. Fulde photographs people. Based in Toronto, ON, he uses bold lighting and vibrant colours to tell people's stories. His work in the film industry lends a cinematic energy to his photographs and makes for an always-colourful studio -- whether he's shooting portraits, fashion, or beauty.

Log in or register to post comments
32 Comments

This is great except do we know what camera brand each poll contributor is actually using? How do we know that Canon users didn't vote for Sony to have the best color, or vice versa. It seems like Tony is just assuming that anyone who voted for Sony to have the best color, must use a Sony camera. Did I miss something? There's a lot of data missing here to call this a "scientific method." If anything it seems like a nice way to troll people for supporting their brand of choice with bias.

Yes you did.
One set of images has false branding on it and other set does not have any branding on it.

Thats just assuming brand loyalty / bias. I don't think he actualy asked what camera they use.

yes

Color Science is the Bro Science of digital photography.

I appreciate that you used the screen grab you did so I could see the results without having to listen to TN blabber on. It tickles me that the highest percentage of his poll is "Don't Care/Doesn't matter."

Those are not real results that he got. Those are results he got when he asked people about the brand not the image. Later he blind tested same people and results were much different

Too bad. Guess I'll see if there's a tl:dw version. I guess I fall into that group anyway.

1. Sony
2. Fuji
3. and 4. Canon and Nikon (or other way around)

Same people voted for branded and blind test.

Canon user after watching the video: ... articulating screen!

You trust the daylight setting of every camera but not the auto white balance setting.
Funny.
Tony’s findings are also confirmed by Max Yuriyev this week. Nikon has a terrible “color science” on their first full frame mirrorless camera. Embarrassing.

Yessssss!!!

“Let the hate run through you”

Now we can move on to, “well your pinky gripability is garbage so I can’t buy that camera”.

Don’t forget articulating screen.

Oh yea! Because that's a make or break capability too. Selfies and youtube are career goals!

Brand wars and jokes aside, I really hope this video opened a few eyes.

I don't know why people care what others think about colors.

If I like Fuji colors and 1000000 people tell me that Fuji colors are crap I would still like Fuji colors.

Saying that color science doesn't matter cause you can fix it in post or that people can't even see it correctly is bullshit. Color science matter a hell of a lot in product photography!

Color science is like aeronautical engineering.

For the people designing the plane, the science is critical. But when a common passenger claims, "I prefer Boeing over Airbus because I can feel their aeronautical engineering is better when take a flight," he's talking out of his fanny.

Same with "color science." If someone prefers the output of one camera to another, that's just a setting preference, not color science.

it was made up to cover the lack of improvement in recent camera's released by large companies.

It seems to me that "color science" are the wrong words for this test. Color preference I believe would have been much better.
I would have liked to see results by camera showing true color comparisons as well as contributors preferences. I would like a camera to produce results as close to true color as possible then I would make changes that please me. The fewer camera color casts added the easier that will be.

Butch

Sony's color is terrible - until you do a blind test

Trying to avoid videos by Tony, FroKnowsEverything and half a dozen other self proclaimed photo gurus but I can't seem to run away from them. Thanks Fstoppers.
One thing I really want to understand, and I hope someone can help me here, why would anyone care about some quiz made up by Tony from SOOC JPEG files? Oh wait. I forgot. Tony does not shoot RAW. My bad.
Well, to everyone else, here is a thought. How about just processing those RAW files, since you will process them anyway, and adjust color to your liking - WB, hue, saturation and luminance for different channels, etc. Yes, it's a bit more work but for a "color scientists" this should be well worth the effort.

Did you watch the video or just spouting off? I can't tell if this is trolling due to how your comment repeats exactly what Tony says in the video to do...

No, this is not trolling but I honestly see no point to spend any more time on this topic. Anyone who has been around long enough knows that Nikon in general has more natural skin tones, and Canon colors are more vivid especially in red channel, and in the end even different cameras for same manufacturer have different color profiles, and all this still means nothing because once you start adjusting ISO weird shit happens to color altogether. Then we have different color profiles, monitors are calibrated in a different way (or not calibrated), same for printers and let's not even start about the actual light under which images are viewed. And when you think you can embrace all this anyway and produce some consistent results just keep in mind recent scientific studies confirm not all people may see same colors alike. Congratulations, you've just accumulated a ton of knowledge which turns out to be highly subjective.
As for the video I honestly do not know what Tony is fighting for. His video has no conclusion and this bothers me probably the most, aside from the fact he is wasting time on it in the first place. If he said "so this is what people think, this is how people voted photos, one poll has nothing in common with the other, color science is a dead end street and bullshit altogether, goodnight" then it would have been fine. But I bet you he will be doing a follow up to this nonsense just to get more clicks on his yt channel.

One of the main points of the video was that people associate color more with brand names than actual... color. What he basically said, is that assumptions like "Anyone who has been around long enough knows that Nikon in general has more natural skin tones, and Canon colors are more vivid especially in red channel" are only true on the internet but not in real world conditions, as long as the camera (brand) is unknown.

First of all, Tony was not asking people which photos reproduce skin tones better or which ones look more vivid. He was asking which photos people LIKE the most. This is a question which is extremely SUBJECTIVE. If you make a poll and don't ask for specifics you are wasting your time because the feedback will not actually represent anything in particular. Tony did exactly such thing.

As for association of colors with brands vs actual preferences I feel this is something only noobs and gear onanists spend time pondering. Because photographers who actually make living taking photos shoot in RAW format and adjust color to their liking be it via profiles or manual adjustments and problem is solved. Any pro photographer endorsed by camera manufacturer who talks about superior color rendition his camera offers is simply full of shit. But there is apparently a significant number of people who can afford A7III or D850 and never go beyond shooting JPEG so this whole color science talk is still a selling point.

People need to understand they are not shooting film anymore where color profile is predetermined by composition of the film emulsion. Digital sensors do not actually register any predetermined colors. This is actually why you can get entirely different colors on SOOC files by adjusting WB. WB adjustment in the camera is not alternating the quality of light of registered scene and yet it produces different results. Isn't it mind-blowing?

This should only come as a surprise to those who aren't paying attention. <-- totally valid situation if you couldn't care less about "color science" (which has been referred to as "color sauce" by some and I think that's hilarious)
Michael the Maven/Mentor on YT compared out of camera jpgs from the A9/1DX Mark II/D5 in a blinded test which later provided the answer key in the same video and the comments section revealed the most people preferred Sony > Nikon > Canon. video: https://youtu.be/0xo9qKPVhEk?t=1753

Then, DPR reposted the video and that comment section revealed the same thing: https://www.dpreview.com/videos/9193994762/blind-portrait-shootout-sony-...
Next, PDN (which Tony references at the end of his video) looked at color reproduction and rated them as follows:
1) A7III
2) A7RIII tied with Fuji X-E3 tied with Leica CL
3) A9
4) Fuji X-H1 tied with Panasonic GH5S
5) D850 tied with Canon M50
6) Pentax K-1 Mark II
7) Canon 6D Mark II
8) Panasonic G9
9) Olympus E-M10 Mark III tied with Fuji X-T100
The caveat to this test is that they perform this test periodically and when they do, they do it with the newest cameras available since the last time they ran the test. So it doesn't rank older camera models and is only a ranking vs the models seen in the list above. But, the take away is that Sony leads the pack, followed by Fuji, followed by Nikon and Canon and Canon's first FF entry came in 7th place as the 11th camera (as there were ties above it).

So, again, this shouldn't come as a surprise.

Damn you did your research +1.

>> All in all, it shows just how little color science matters. The difference is minute. <<

No, it doesn't show that. What your comment shows is that you are an overconfident ignoramus: colour science is not about defined by the getting the awb right for jpegs.

To get an idea how much more -

https://www.provideocoalition.com/camera-not-color-gamut/

https://www.provideocoalition.com/camera-may-color-gamut-musings-camera-...

"...I found almost nobody considers themselves an expert"

Clearly he didn't poll the FStoppers comment section. ;)

Real world story example!
I shoot Canon. My friend shoots Sony. Both cameras preformed wonderful. Colors looked great. But then we shot this video in a badly lit warehouse. The lighting was a mix of all different types of lights ( incandescent, florescent, halogen, vapor). My Canon did really well. Skin tones held up in most shots. My friends Sony..... It crashed and burned. He was so upset that he couldn't get good skin tones even shooting the same kelvin as I was. He sold his Sony after a few other projects gave him trouble.
Moral of the story. Your cellphone can take as good a photo as my Canon 1dx mark ii (almost), but it has to be in certain setups and scenarios (aka perfect situations). This has kept me switching. Real world experiences are more important.